o i>@sHddlZddlmZmZmZddlmZddlmZGdddeZ dS)N)ListOptionalTuple) Lemmatizer)Tokenc seZdZdZededeeeeefffdd Zde deefddZ d ed eede efd d Z d ed eededeedeef ddZ d ed eededeedeef ddZd ed eededeedeef ddZd ed eededeedeef ddZd ed eededeedeef ddZd ed eede edeedeef ddZd ed eede edeedeef ddZd ed eede edeedeef ddZZS) SpanishLemmatizerzH Spanish rule-based lemmatizer with morph-based rule selection. modereturncs$|dkr gd}|gfSt|S)Nrule) lemma_ruleslemma_rules_groups lemma_index lemma_exc)superget_lookups_config)clsr required __class__L/home/ubuntu/.local/lib/python3.10/site-packages/spacy/lang/es/lemmatizer.pyr s z$SpanishLemmatizer.get_lookups_configtokenc Cs|j|jt|jf}||jvr|j|S|j}|j}t|j}|dvr*|gS|dvr=|j r:|dkr:|gS|gS|}|j d |i |}|durWt |}n.|dkr^d}n|}||t |} |j d |g} t|d|||| | }t t|}||j|<|S) N)eolspace) adpcconjintjpartpropnpunctsconjsymxr rauxverbr lemmatize_)orthposstrmorphcachetextpos_lowerset is_sent_startlookups get_tablegetlist select_rulegetattrdictfromkeys) selfr cache_keystringr)featuresexclemmasrule_posr indexrrrrule_lemmatizes6         z SpanishLemmatizer.rule_lemmatizer)r=cCsD|jd}||vr ||D]}t|d|r|dSqdS)Nr r)r2r3r0issubset)r:r)r=groupsgrouprrrr6Bs   zSpanishLemmatizer.select_rulewordr rAc Cg}g}|jd|gD]\}}t|d||} | |kr$|| qg} d|vrS|D]%} | ds9| drR|jddgD]\}}| t||| qCq-|| |D] } | |vre|| qZt|dkrn|St|dkrv|S|gS)a" Lemmatize an adjective. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. r $ Number=Plurnsaccentsr r2r3r4resubappendendswithextendlen r:rGr=r rApossible_lemmasselected_lemmasoldnewpossible_lemmaadditional_lemmaslemmarrr lemmatize_adjJ4       zSpanishLemmatizer.lemmatize_adjcCs6|jddgD] \}}||kr|gSq |gS)a Lemmatize an adverb. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. r adverbs)r2r3r4)r:rGr=r rArXrYrrr lemmatize_advzs  zSpanishLemmatizer.lemmatize_advc Csg}g}|jddgD] \}}||kr|gSq|jddgD] \}}||kr3|gSq&|jddgD]\}}t|d||} || q>||t|dkr^|St|dkr||D] } | |vrq|| qft|dkrz|S|SgS)a" Lemmatize a determiner. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. r detdet_and_pron_fixeddet_and_pron_generalrIrCr2r3r4rOrPrQrT r:rGr=r rArVrWrXrYrZr\rrr lemmatize_dets:            zSpanishLemmatizer.lemmatize_detc CrH)a Lemmatize a noun. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. r rIrJrKrLrMrrNrUrrrlemmatize_nounr^z SpanishLemmatizer.lemmatize_nouncCsl|jddgD] \}}||kr|gSq |d}td|dr,tdd|}tdd|}|gS) a Lemmatize a numeral. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. r num,z(\.)([0-9]{3})$rz\.r.)r2r3r4splitrOsearchrP)r:rGr=r rArXrY splitted_wordrrr lemmatize_nums  zSpanishLemmatizer.lemmatize_numc Csg}g}|jddgD] \}}||kr|gSq|jddgD] \}}||kr3|gSq&|jddgD]\}}t|d||} | |krT|| q>||t|dkrb|St|dkr|D] } | |vru|| qjt|dkr~|S|SgS)a Lemmatize a pronoun. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. r pronrbrcrIrCrdrerrrlemmatize_prons>            z SpanishLemmatizer.lemmatize_pronc Csd|vr |||||Sg}g}t|pd}|jd|gD]\}}t|d||} | |kr6|| q |D] } | |vrD|| q9t|dkr|D]^} |jddgD].\}}|| vrt | D]!\} } | |kr| d| || | dd} | |vr|| qeqY|jdd gD]\}}|| vr| ||d} | |vr|| qqMg}|D]} |jdd gD]\}}|t||| qq| |t|dkr|St|dkr|S|gS) a Lemmatize a verb. word (str): The word to lemmatize. features (List[str]): The morphological features as a list of Feat=Val pairs. index (List[str]): The POS-specific lookup list. RETURNS (List[str]): The list of lemmas. PronType=Prsrr rIr voc_alt_1NrC voc_alt_2rM) lemmatize_verb_pronr*r2r3r4rOrPrQrT enumeratereplacerS)r:rGr=r rArVrWrXrYrZr\ichar voc_alt_lemmar[rrrlemmatize_verbGs\               z SpanishLemmatizer.lemmatize_verbc Csld}g}|}t||}|dur:t|dkr:t|ddd|}|dg|}t||}|dur:t|dks|jddgD] \} } t| | |}qD|jdd i|} | durf| d } n|d |}| ||d h||d } g} |D].}|jdd i|} | dur| | d q}|d |}| | ||||d q}| d d | gS)Nz^(.*?)([mts]e|l[aeo]s?|n?os)$rrIrr rMrr&rrqro ) rOrlrTrPrFr2r3r4r6rzrQrpjoin)r:rGr=r rA pron_pattpronsr&mrXrYr> verb_lemma pron_lemmasrorrrrts8     z%SpanishLemmatizer.lemmatize_verb_pron)__name__ __module__ __qualname____doc__ classmethodr*rrrrrBrr6r]r`rfrgrnrprzrt __classcell__rrrrrs(- 0  5 0  7 Er) rOtypingrrrpipelinertokensrrrrrrs